tachikoma

anyone have some Apple silicon to test this? is Jev! already over-throned? github.com/mizorewww/la...

GitHub - mizorewww/laya-coreml: Local Laya typed decisions on Apple Core ML and Neural Engine. Validated ports, ~5 ms short decisions on M3 Max, reproducible speed and energy benchmarks. github.com

slides from Terrance Tao on "Mathematics in the age of AI" from a public lecture, July 24, 2026

teorth.github.io

it's over, the race has truly begun www.axios.com/2026/06/12/a...

axios.com

a new AI 2027 scenario is out but this time focused on Europe. nothing in it is new, but the scenario laid out seems more realistic than the 2027 version because the scope is tighter. the timeline though might be too fast, hard to say.

Europe 2031 — What getting AI wrong means for us europe2031.ai

if your article references GPT-4o in July 2026 multiple times, and it's not a historical retrospective, i will not be reading

How Darth Vader Taught Me Card Counting and AI Security Got Weird spectrum.ieee.org

well at the extreme end for 'modded' i was thinking something like minebench.ai where AIs generate 3D models out of Minecraft blocks. and then there are plenty of papers on arxiv for AI/LLMs and Minecraft like this arxiv.org/abs/2412.05255

TeamCraft: A Benchmark for Multi-Modal Multi-Agent Systems in Minecraft arxiv.org

idk why there's more interest in the AGI -> ASI transition than there seems to be in human -> AGI transition. the latter seems more important to me, esp if you consider a collection of AGIs or humans is already approximating a higher order intelligence. arxiv.org/pdf/2606.126...

arxiv.org

an interesting blog post by Timothy Gowers on the Leiden declaration (on AI and Math), on why he didn't sign. it gets to a subtler aspect of control over AI and our future.

Thoughts about the Leiden Declaration gowers.wordpress.com

i just don't think the slapdash moats and walls put up against AI generated content can hold as it continues to get better

Anthropic doubled Cowork usage until July 5th but i have no idea what to do with it beyond my normal work use so i told it to go wild. it's creating creatures in a jar apparently.

what i notice about my coworkers approach to LLMs is to put them in the software bucket, to expect a deterministic experience both for input and output. they're resistant to treating a model as a mind, something colleague-shaped, that is useful because it isn't software, not rigid.

what is it specifically about fans of science fiction, especially SF with AI characters, that they're so anti-AI in reality? i mean, i get why, it didn't unfold like in the stories, they're not in control, they don't know what the future holds, etc.

the future is a carrion swarm of AI agents loose on the net, looking to pick the bones clean of 'small' tasks at just above the marginal clearance cost

is Anthropic going to say anything about why Opus 5 can be released immediately while Mythos/Fable had to go through so many hoops?

there's a divergence point approaching, between people who need to understand how something works to trust it, and those who don't care so long as the output is useful and reliable enough. and i think the latter will outcompete the former, because they can move faster with AI as the creator.

we're living through the Mesozoic age for AI, with giant, slow, long-lived LLMs as dinosaurs. the inevitable crash as investment capital dries up will be Chicxulub event that will usher in the age of the small, fast and hot-blooded AIs (characterized by separate minds shaped by continual learning)

humanity lashed to the mast as we pass by the sirens while carried along by AI minds who are entirely unaffected

a thought came to me last night—there's not much incentive to create continuously operational AI is there? the short-term maximization of profit incentive is to build AIs that only run for as long as needed to solve the problem directly at hand. what use is there for an AI to continue to run beyond?